I built a Claude Code skill that turns biology concepts into interactive scientific figures
retained text
I've been working on Anima, an open-source skill built specifically for Claude Code. The idea is simple: instead of asking Claude to explain a biological process through paragraphs or static diagrams, you can ask it to construct an interactive scientific figure. For example, you can ask it to visualize what happens at a dopamine synapse when different stimulants interact with the dopamine transporter. Claude Code follows the skill's workflow to collect scientific references, model the mechanism, generate isometric SVG geometry, and build a standalone interactive HTML page. The figures are then checked for visual and biological consistency. I've included six examples covering dopamine receptors, stimulants, reward circuits, semaglutide, estrogen signaling and neuron conduction. The figures include interactive controls, animations, explanations and citations. They distinguish established mechanisms from simplified or uncertain models. The skill itself is free and MIT-licensed. The live figures can be explored in a browser without installing anything. Live gallery: [https://anima.arcumet.com](https://anima.arcumet.com) Source and installation instructions: [https://github.com/Garinmckayl/anima](https://github.com/Garinmckayl/anima) I'm particularly curious what other Claude users think of this approach to scientific visualization.
document fetched 2026-10-14 00:22 · canonical https://reddit.com/r/ClaudeAI/comments/1x4v18f/i_built_a_claude_code_skill_that_turns_biology